Comprehensive Analysis of Display Advertising
Analysing display ads isn’t easy. Firstly, a lot of people don’t understand how display advertising affects conversions and consider it an unnecessary luxury. Second, display ads tend to run across all available channels, and even the simple aggregation of data from Facebook, Google, and direct sites becomes a major challenge.
In this article, newage. will share an approach to display ad analytics. We will show you how to collect the right data and how to analyse the real effectiveness of campaigns.
Comprehensive analysis of display advertising is a methodology for evaluating the effectiveness of display advertising campaigns. As part of it, we analyse the quality of placement, media metrics, and reaction to advertising, which can be measured using digital tools.
How We Use the Comprehensive Analysis
Comprehensive analysis is the main approach of newage.agency to display advertising. It refers to an iterative cycle through which we optimize campaigns. And this iteration consists of 4 parts.
1. Quality control of advertising placement. Before examining any data, you need to make sure that it is correct and reflects the real course of the campaign. How are things going with viewability, was the advertising contact visible by the user?
2. Display metrics analysis. Display advertising is a special type of promotion, which has its own basic performance indicators: impressions, target audience reach, video views, brand lift, etc. We measure them at this stage.
3. Reaction to advertising. Based on the analysis of data about post-click, post-view, and cross-device conversions, conclusions regarding how the user was influenced by interactions with the advertising could be derived. In fact, this is performance marketing, but for brand campaigns — brandformance.
4. Agile approach. Our task is to draw as many conclusions as possible in the shortest possible time, in order to use these findings for optimization. Our approach allows us to invest your marketing budget with maximum return.
Let’s take a closer look at each stage.
Quality Control of Advertising Placement
Incorrect data leads to erroneous conclusions and disastrous actions. Therefore, the first and fundamental step in comprehensive analysis is data validation.
Display ads are launched across many channels at once. We simultaneously use advertising offices of Google, Facebook, and other global and local players and direct purchases.
In order to keep track of all the placements, we additionally set up a tracking system (auditor), which helps us to compare all indicators. We aggregate statistics from advertising offices, auditors, and sites in a single dashboard and check if it converges.
First, we verify the received data – whether the indicators of advertising cabinets, web analytics, and auditors converge. After all, it is wrong to draw conclusions if you have a million impressions in your ad account for which you paid, and the tracking system displays only 100 thousand impressions.
Then we check the quality indicators of the placement itself. We find out whether the user could see the advertisement, whether the planned placement format matches the actual one, on which site, and how often the audience sees the advertisement.
The viewability indicator (active view) is especially important here. If the viewability of your placements is, say, 10%, then it would be incorrect to speak of 100% coverage, which is traditionally shown as total reach.
It is important to check these parameters at the stage of quality control of placement.
Earlier, we got into situations where the data was tracked incorrectly. As a result, we are acting proactively in order to prevent this from happening.
In favour of developed check-lists and results of the pretest derived from 1000 impressions before the main launch, we are able to proactively prevent data tracked incorrectly. These verification methods work, and over the past few years, such situations have become rare exceptions.
Display Metrics Analysis
In this section, we analyse the relatively basic media indicators:
- impressions;
- target audience coverage;
- coverage of the target audience at the frequency;
- inspections;
- brand lift;
- baseline brand health indicators;
- growth of brand inquiries;
- growth of direct traffic, etc.
An important aspect of this part is the coverage assessment only on active view (viewability) impressions. Unfortunately, few people on the market follow this, and this is an important point.
Many media metrics are based on research where a focus group meets and is interviewed about a brand, product, or campaign. This is a standard method of research, but it has much larger errors than in logged, technometric indicators.
These indicators are also important to look at and analyse, and should not be overlooked in the Comprehensive Analysis. They will help supplement the general conclusions on campaigns both in the context of different media and for clients who find it difficult to track the reaction to advertising in digital.
Reaction to the Ads
In display advertising, the user does not immediately follow the link to order the product. There are several reasons for this.
- The media format “catches” the user at an unfortunate moment. For example, when he is going to watch interviews on YouTube instead of ordering diapers of a certain brand.
- Viewers don’t remember display ads the first time. It is required to show the video several times in order for the user to become loyal enough to switch.
- Long sales cycle products require the customer to be warmed up before purchasing. You can’t just add an apartment to the cart just by clicking on a branded video.
Without numbers, all this sounds like some kind of magic, but this is a very real and measurable effect – the media effect of advertising. To get the coveted numbers and link display ads to performance indicators, we analyse the delayed user actions after viewing ads. And there are three types of conversions here.
1. Post-clicks are transitions to a target site directly from an advertisement. The user saw the ad and immediately clicked.
2. Post-view conversion is the execution of targeted actions a while after viewing an advertisement. Yesterday I saw an advertisement, and did not click, but today I went to the site for a branded request. The tracking system will connect this view and search.
3. Cross-device conversion is action delayed in time and from other devices. For example, a viewer saw a video in the subway on a smartphone, and the next day found a company from a laptop.
Most advertisers analyse only post-click conversions because this indicator is available in any ad account. But this is a small piece of the media effect.
Our experience shows that of those who went to the site after viewing ads, clicks from ads are just 20-30%. And among those who, after the campaign, reached the conversion, there are less than 10% of them.
If you don’t have data about post-view and cross-device conversions, then it is better not to look at clicks at all. From incomplete data (20-30% of reactions), you are guaranteed to draw the wrong conclusions. In this case, it is better to analyse only media indicators.
Using post-click + post-view + cross-device data, you can measure the user’s reaction, and most importantly, draw conclusions that will help optimize your campaign:
- What is the optimal frequency for the campaign?
- Which creative is effective and which is not?
- How often should users see an advertisement, for how long will they remember it?
- Which sites/targeting works and which does not?
- Through which channel (search, direct, advertising) does the user come to the client’s site, after contacting the display advertising?
Agile Optimization
Comprehensive analysis of display advertising is an integrated, cyclical approach. It is not needed in order to check the campaign one time and say: “What cool guys, they did everything well.“ The most important component of any campaign is its additional optimization, constant testing of hypotheses, and squeezing the maximum out of the budget.
It is important to repeat all the previous stages of the analysis on a regular basis in order to optimize the campaign. We have several times estimated what would have happened if the campaign had continued to spin with the initial settings.
And it turned out that each iteration of the edits increases the efficiency of using the budget from 20-30%, and the overall efficiency relative to the start in such tests grows from 100%.